Nanonets-OCR vs MiniCPM-V 4.5
Pick Nanonets-OCR for printed business documents that you want as Markdown, with signatures, watermarks, checkboxes and flowcharts tagged in the output. Pick MiniCPM-V 4.5 if you need a clear Apache-2.0 licence, quantized builds for edge or single-GPU use, or document Q&A with free-form prompts. The main split is that Nanonets-OCR is tuned to one fixed prompt and one output style, while MiniCPM-V 4.5 is a general vision language model you can prompt freely.
Nanonets-OCRnanonets/nanonets-ocr | MiniCPM-V 4.5openbmb/minicpm-v | |
|---|---|---|
| Type | Open source | Vision model |
| Price | 2 credits / page | 2 credits / page |
| Per 1,000 pages | $2.00 | $2.00 |
| Typical speed | 1.8 s / page | 2.5 s / page |
| Best for | Markdown with semantic tags | OCR on edge devices |
| Inputs | PDF, images | images |
| Languages | Multilingual | 30+ languages |
| Max pages per request | 500 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | ✓ Yes |
| Math | ✓ Yes | No |
| Bounding boxes | No | No |
| Confidence | No | No |
| Forms | ✓ Yes | No |
| Multilingual | ✓ Yes | ✓ Yes |
| On OCRRouter | Available | Available |
Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.
Which to choose
Choose Nanonets-OCR if
- It tags signatures, watermarks and image descriptions, and turns checkboxes into unicode symbols.
- It extracts flowcharts and org charts as Mermaid code and tells inline LaTeX from display LaTeX.
- It is a 3B model with a published DocVQA score of 89.43 and an olmOCR-bench overall score of 69.5 on its model card.
Choose MiniCPM-V 4.5 if
- Its weights and code are Apache-2.0, while the Nanonets-OCR2-3B model card states no licence.
- It runs on llama.cpp and vLLM, with int4, GGUF and AWQ builds for on-device or single-GPU use.
- It accepts free-form prompts and can return plain text, Markdown or JSON, so it fits local document Q&A.
Key differences
| Nanonets-OCR | MiniCPM-V 4.5 | |
|---|---|---|
| Licence | Not stated on the Nanonets-OCR2-3B model card; base model is Qwen2.5-VL-3B | Apache-2.0 |
| Model size | 3B (Hugging Face lists about 4B params) | 8.7B, built on Qwen3-8B and SigLIP2-400M |
| Output formats | Markdown, HTML, LaTeX, Mermaid | Plain text, Markdown, JSON |
| Prompting | Trained on a fixed prompt; other prompts are expected to hallucinate | General vision language model; no such limit listed |
| Handwriting | Not trained on handwriting and can hallucinate; one test saw generation loops with CER up to 2100% | Not published |
| Published benchmarks | olmOCR-bench 69.5, DocVQA 89.43, ChartQA 78.56 (OCR2-3B, by the authors) | Not published as text; the model card gives scores only as images, though OpenBMB claims leading OCRBench and OmniDocBench results |
| Runtime support | Self-host from Hugging Face; an open Ollama issue exists | llama.cpp, vLLM, Transformers; Ollama and SGLang were still pending on the model card |
Sources
- Nanonets-OCR2-3B model card
- Nanonets-OCR-s announcement
- HF discussion fixed prompt
- Ollama issue #11100
- Hugging Face model card
- MiniCPM-V 4.5 paper
What each one is
Nanonets-OCR
Open image-to-Markdown model that tags tables, equations, signatures, watermarks and checkboxes.
Nanonets-OCR detailsMiniCPM-V 4.5
8B vision model from OpenBMB (August 2025) built to run on devices, with strong OCR and document parsing.
MiniCPM-V 4.5 detailsTry both on your own file
Run the file once with each model and compare the output.
Use one, fall back to the other
{"model": "nanonets/nanonets-ocr","fallbacks": ["openbmb/minicpm-v"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Nanonets-OCR alternativesMiniCPM-V 4.5 alternativesBest OCR models in 2026